A deep learning based wearable system for food and drink intake recognition
- PDF / 2,069,084 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 86 Downloads / 165 Views
ORIGINAL RESEARCH
A deep learning based wearable system for food and drink intake recognition Dario Ortega Anderez1 · Ahmad Lotfi1 · Amir Pourabdollah1 Received: 1 May 2020 / Accepted: 6 November 2020 © The Author(s) 2020
Abstract Eating difficulties and the subsequent need for eating assistance are a prevalent issue within the elderly population. Besides, a poor diet is considered a confounding factor for developing chronic diseases and functional limitations. Driven by the above issues, this paper proposes a wrist-worn tri-axial accelerometer based food and drink intake recognition system. First, an adaptive segmentation technique is employed to identify potential eating and drinking gestures from the continuous accelerometer readings. A posteriori, a study upon the use of Convolutional Neural Networks for the recognition of eating and drinking gestures is carried out. This includes the employment of three time series to image encoding frameworks, namely the signal spectrogram, the Markov Transition Field and the Gramian Angular Field, as well as the development of various multi-input multi-domain networks. The recognition of the gestures is then tackled as a 3-class classification problem (‘Eat’, ‘Drink’ and ‘Null’), where the ‘Null’ class is composed of all the irrelevant gestures included in the post-segmentation gesture set. An average per-class classification accuracy of 97.10% was achieved by the proposed system. When compared to similar work, such accurate classification performance signifies a great contribution to the field of assisted living. Keywords Gesture recognition · Accelerometer · Deep learning
1 Introduction Eating difficulties are those that alone or combined, hamper the preparation or the intake of food and/or beverages (Westergren 2001), with major causes including cognitive impairment, poor appetite or feeding dependency. Incidentally, a poor diet can contribute to weight loss and malnutrition, leading to potential functional limitations, metabolic abnormalities and diminished immunity (Payette and Shatenstein 2005). Recent statistics outline eating difficulties as a prevalent issue among the elderly population. For instance, the survey conducted in Westergren et al. (2002) with 520 elderly patients in hospital rehabilitation, reveals 82% of * Dario Ortega Anderez [email protected] Ahmad Lotfi [email protected] Amir Pourabdollah [email protected] 1
School of Science and Technology, Nottingham Trent University, Clifton Campus, College Drive, Nottingham NG11 8NS, UK
the patients exhibit some form of eating difficulty. The survey conducted in Lohrmann et al. (2003), including 3000 patients from 11 different hospitals, acknowledge 21.1% of the patients younger than 80, and 36.4% of those aged 80 or older require eating assistance. As of now, dietary behaviour is generally monitored by the use of self-assessment questionnaires. Though, two major shortcomings are found on the use of these conventional approaches. First, the data entry process may result
Data Loading...